We are interested in establishing the physiological role of well-established neurochemical pathways (such as NO and COX-2) in shaping the spatial specificity of CBF regulation. The working hypothesis is that increases in CBF are mediated by vasoactive substances released by strategically located cells in the parenchyma that act on the cerebral vessels and regulate their tone. The outstanding questions are: (i) which substances are released by what cells and under which circumstances? (ii) What is the spatial specificity of with respect to the cortical architecture? (iii) What is the physiological role of vasoactive agents in the context of underlying physiological and pathophysiological processes? The main approach has been to use pharmacological manipulations to determine the relative contribution, spatial specificity and cellular basis of the different pathways to the HRF. Future experiments will focus on understanding the relevance of these pathways to neurovascular coupling in the presence of pathophysiological states such as hypertension and ischemic stroke. In the current review cycle, we set out to investigate in further details the specific role of COX-1, COX-2 and their vasoactive agents in neurovascular coupling. To target the isoforms COX-1 and COX-2, but also the endothelial PGE2 receptors EP1 and EP2, the cranial window was superfused with selective inhibitors and antagonists, and the functional responses during pharmacological inhibition were compared with those obtained during superfusion of the cranial window with vehicle. Seven experiments were carried out under different pharmacological conditions. Experiments 1 and 2 tested the role of COX-2 in neurovascular coupling. We found that cortical superfusion of the somatosensory cortex with the specific COX-2 inhibitor NS-398 (6x10-4M) caused a significant attenuation of the CBF response to contralateral forepaw stimulation to 60% of the physiological response obtained during superfusion with vehicle. Furthermore, superfusion with PGE2 (10-5M), one of the vasodilatatory products of cyclooxygenase, resulted in recovery of the NS-398 attenuated CBF response back to 85% (P <0.05 versus NS-398) of the original response. However, superfusion of PGE2 in the absence of prior COX-2 inhibition (Expt. 3) did not enhance the response to somatosensory stimulation. Experiments 4 and 5 tested the role of COX-1 in neurovascular coupling. In contrast to the attenuating effect of COX-2 inhibition, superfusion with the preferential COX-1 inhibitor SC-560 (5x10-4M, Expt. 4) did not decrease the CBF response. Furthermore, SC-560 showed no additive effects to the attenuation of the CBF response caused by NS-398 (Expt. 5, P >0.05). Experiments 6 and 7 tested the role of the PGE2 receptors EP1 and EP4 in neurovascular coupling. Local superfusion of the cortex with SC-19220 (10-4M), a selective EP1 antagonist, in contrast to EP4 inhibition with the selective antagonist GW-627368X (2x10-4M), resulted in significant attenuation of the CBF response to 75.5% of the vehicle response. Combined superfusion of the responsive area with a mixture of EP1 and EP4 antagonists did not show an additive effect (Expt. 7). In agreement with our previous study, there was no significant effect of any of the pharmacological treatments on either amplitude or latency of the SEP, suggesting that inhibition of either isoform of cyclooxygenase or PGE2 receptors has no effect on the evoked local neuronal activity. We then used 2PLSM to acquire stacks of 80-90 images parallel to the surface of the somatosensory cortex and covering up to 450 &#956;m below the pial surface. For each rat, three sets of two image stacks were acquired before and during sustained somatosensory stimulation, during superfusion of the cranial window with vehicle, with NS-398 and with PGE2. Three subsets of images were chosen at the level of the pial surface, layer I and layers II-III, co-registered with the same images in the other stacks, and compared across different conditions. During superfusion of vehicle, forepaw stimulation produced an increase in the area covered by dextran-rhodamine by 4.4 3.0%, 28.4 15.5% and 13.4 5.2%, at the levels of pial vessels, layers I and II-III, respectively. Superfusion of the cortex with the COX-2 inhibitor NS-398 attenuated the vasodilatatory response of the microcirculation. Subsequent superfusion with PGE2 recovered the attenuated microvascular response in the pial surface and in layer I, but not in layers II-III. Quantification of the data shows a significant increase in the number of c-Fos immunopositive cells in the contralateral somatosensory cortex, which co-localize with COX-2 and with the pyramidal cell marker PYR positive cells, but do not co-localize with COX-1 positive cells. Further neurochemical identification of the COX-1 and COX-2-positive cells by double immunofluorescence reveals that the vast majority of the investigated COX-2 cells are also positive for the pyramidal cell marker, confirming that COX-2 cells are subpopulation of PYR neurons. In contrast, COX-1 cells co-localize neither with the rat brain pyramidal cell marker, nor with the astrocytic marker S100&#946;, nor with COX-2, but only with the microglial cell marker Iba1, confirming that the only physiological cortical origin of COX-1 is in microglia. This goes against data presented in recent reports of robust localization of COX-1 in astrocytes, and leads us to conclude that the reports that have implicated COX-1 as a main mediator of neurovascular coupling were probably looking at secondary pathways that were upregulated in response to a neuroinflammatory reaction of the tissue either to the surgery to equip the animal with a cranial window, or to the chemicals used to visualize calcium waves in the brain, or even to the pulses of laser light used to uncage Ca2+ next to the astrocytes. All of these procedures may put the cortical tissue under non-physiological conditions, prone to yield misleading results. It will be interesting to compare in further detail the cortical expression of COX-2 with the laminar profiles of the CBF response to increased neural activity. The Role of NO: Due to its major function as a potent vasodilator, nitric oxide (NO) has been frequently interrogated as the most prominent signaling molecule in neurovascular coupling. In the current review cycle, we extended our previous work to investigate the effectiveness of two NO-releasing molecules the direct NO donor sodium nitroprusside (SNP) and the metabolic NO precursor nitrite, as sources of NO in neurovascular coupling. Nitrite is a therapeutically promising vasodilatatory compound, as it is not only metabolized into NO, but also alters ATP release from erythrocytes and can potentially increase H2O2 levels. Working in collaboration with the lab of Dr. Alan Schechter in NIDDK, we showed that systemic inhibition of nNOS with the selective inhibitor 7-NI decreased both the CBF and the BOLD fMRI responses to somatosensory stimulation in &#945;-chloralose anesthetized rats without affecting the resting CBF baseline, as previously reported. In addition, 7-NI also attenuated the CBF response measured in rats equipped with a cranial window. Systemic IV infusion of the NO-donor SNP to achieve a vascular concentration of 10nM significantly recovered the BOLD response, but not the CBF fMRI response. When evaluated by LDF, however, topical superfusion of 10nM SNP through the cranial window significantly recovered the functional response. Interestingly, superfusion of the cranial window with a 1 &#956;M nitrite solution fully restored the neurovascular coupling to the pre-drug levels. The data shows that systemic administration of NO can partially restore the hemodynamic response to somatosensory stimulation, in agreement with previously published results.